CN111951116A - Medical insurance anti-fraud monitoring and analyzing method and system based on unsupervised isolated point detection - Google Patents
Medical insurance anti-fraud monitoring and analyzing method and system based on unsupervised isolated point detection Download PDFInfo
- Publication number
- CN111951116A CN111951116A CN202010868057.9A CN202010868057A CN111951116A CN 111951116 A CN111951116 A CN 111951116A CN 202010868057 A CN202010868057 A CN 202010868057A CN 111951116 A CN111951116 A CN 111951116A
- Authority
- CN
- China
- Prior art keywords
- isolated
- medical insurance
- sample
- insured person
- characteristic data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Quality & Reliability (AREA)
- General Engineering & Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Operations Research (AREA)
- Artificial Intelligence (AREA)
- Tourism & Hospitality (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Technology Law (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention provides a medical insurance anti-fraud monitoring and analyzing method based on unsupervised isolated point detection, which comprises the following steps: step S1, acquiring the visit behavior characteristic data of the ginseng insurance people to form an original sample set; step S2, obtaining a plurality of visiting behavior characteristic data sets of the insured persons according to the original sample set; step S3, creating a corresponding isolated tree according to the visit behavior feature data set of each insured person; step S4, integrating the isolated trees to form an isolated forest; step S5, establishing an evaluation index for evaluating the visit behavior of the insured person based on the isolated forest; and step S6, traversing each isolated tree of the isolated forest for the input visit behavior characteristic data sample of the insured person, and outputting a detection result according to the obtained evaluation index. The invention improves the detection level of the fraud behavior of the insured person and improves the working efficiency of the mechanism.
Description
Technical Field
The invention relates to the field of medical insurance anti-fraud monitoring, in particular to a medical insurance anti-fraud monitoring and analyzing method based on unsupervised isolated point detection.
Background
Medical insurance fraud is the act of deceiving medical insurance treatment or medical insurance funds by citizens, law enforcement or other organizations violating medical insurance management laws and policies, making false positives, concealing actual conditions, etc. The frequent occurrence of medical insurance fund fraud is a great hazard to the operation and development of medical insurance systems. The medical insurance loss is huge, and the problem to be solved is urgent in regions.
The medical insurance system involves many stakeholders, much more complex than other risk categories, mainly including medical suppliers (hospitals, pharmacies, etc.), medical demanders (paramedics), medical insurance management departments (offices and supervising agencies). Because the related links are more, the chains are long, and the risk points are more, if the measures of the fund risk of the method are not completely taken, the problems of medical insurance fraud and cheating insurance are easily bred. Current subjects of medical insurance fraud can be divided into medical institution fraud, medical insurance participation fraud (insurer fraud) and doctor-patient conspiracy fraud.
The medical database contains reliable and transparent standard medical information, and in the big data era, all medical related data including diagnosis and treatment information, treatment records, personal basic information, participation situations, medicine and instrument use information and the like of patients are stored. An effective medical insurance fraud early warning model is established by utilizing massive medical data, and decision support is provided for the medical insurance center to implement supervision work, so that the method is the primary task to be solved at present.
The medical insurance anti-fraud inspection not only needs a large amount of manpower, material resources, financial resources and time, but also needs continuous technical support. The current technology for preventing fraud of medical insurance in China still depends on blacklist and rule base to monitor and manage the safety of medical insurance fund. The blacklist is an anti-fraud technical method which is used by medical insurance supervision and management departments to check out fraud behaviors and bring the fraud behaviors into the medical insurance fraud blacklist. However, this method has two fatal defects, one is postfix and must be put in a blacklist after being checked by a monitoring person, and an early warning mechanism is lacked. Another is the lack of dynamic management, i.e. cases that go into the black list, will be permanently on the black list, and when the case behaves normally, lack of mobility management, it will be pulled out of the black list.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, provides a medical insurance fraud prevention monitoring and analyzing method and system based on unsupervised isolated point detection, can help relevant organizations to better identify medical insurance fraud behaviors of participants and insurers, and improves the generalization capability and accuracy of the system by adopting an integrated idea.
The invention provides a medical insurance anti-fraud monitoring and analyzing method based on unsupervised isolated point detection, which comprises the following steps:
step S1, acquiring the visit behavior characteristic data of the ginseng insurance people to form an original sample set;
step S2, obtaining a plurality of visiting behavior characteristic data sets of the insured persons according to the original sample set;
step S3, creating a corresponding isolated tree according to the visit behavior feature data set of each insured person;
step S4, integrating the isolated trees to form an isolated forest;
step S5, establishing an evaluation index for evaluating the visit behavior of the insured person based on the isolated forest;
and step S6, traversing each isolated tree of the isolated forest for the input visit behavior characteristic data sample of the insured person, and outputting a detection result according to the obtained evaluation index.
Further, step S3 specifically includes:
sample data set X ═ X1，x2……xnThe visit behavior characteristic data set of the insured person is obtained; wherein
xi＝(xi1,xi2,……，xid)；
S301, randomly extracting from X
A subset X' of X formed by sample points is placed into a root node;
s302, randomly appointing a dimension q from d dimensions, and randomly generating a segmentation point p, min (x) in the current subsetij,j＝q,xij∈X’)＜p＜max(xij,j＝q,xij∈X’)；
S303, the dividing point p generates a hyperplane, and the current data space is divided into two subspaces: assigning sample points with the corresponding value of dimension smaller than p to be classified into a left child node, and assigning sample points with the value larger than or equal to p to be classified into a right child node;
s304, recursion S302 and S303, until all leaf nodes have only one sample point or the isolated tree has reached a specified height.
Further, the evaluation index is established as follows:
for each sample xiTraversing each isolated tree of the isolated forest, and calculating the average path length of the isolated trees in the isolated forest
E (h (x)) is sample xiPath length expectation in solitary forests; average path length for all samples
Carrying out normalization processing; the calculation formula of the evaluation index is as follows:
wherein,
h (i) is a harmonic number.
Further, in step S6, an evaluation index threshold is specifically set, and when the obtained evaluation index is greater than the evaluation index threshold, it is determined that the behavior of the insured person is abnormal; the corresponding ginseng protector is a suspected ginseng protector.
Further, step S2 specifically includes:
for an original sample set, when the number of samples is larger than a sample number threshold value, simply dividing the original sample set into a plurality of visit behavior characteristic data sets of the insured persons;
when the number of samples does not exceed the threshold value of the number of samples, extracting the samples from the original sample set for multiple times by adopting a bagging method to obtain k diagnosis behavior characteristic data sets of the insured person; k is more than or equal to 2.
In a second aspect of the present invention, a medical insurance anti-fraud monitoring and analyzing system based on unsupervised isolated point detection is provided, which includes:
a memory storing a computer program;
a processor for executing the computer program, the computer program when executed performing the steps of the method as described above.
The invention has the advantages that:
1) the detection level of the fraud behaviors of the insured person is improved;
the anti-fraud policy and mechanism perfection of medical insurance are in the process of routing and building dense drums, but the behaviors of the anti-fraud of medical insurance and the loss of medical insurance funds are still high, and the anti-fraud monitoring and analyzing system of medical insurance based on unsupervised isolated point detection adopts an artificial intelligence algorithm to deeply combine with the fraud behavior data of the insured person, so that the capability upgrade of an anti-fraud system is realized; the normal behavior of the insured person can be brought into the model, the detection and detection fraud behavior of the insured person is expanded, the medical insurance fraud occurrence rate of the insured person is effectively prevented, and the anti-fraud effect is obvious.
2) The resource cost is saved;
the traditional fraud detection method needs to consume a large amount of manpower, financial resources and time to identify and judge the fraud behavior, and is particularly insufficient in the presence of huge data volume and rich fraud means; the medical insurance anti-fraud monitoring and analyzing system based on unsupervised isolated point detection reduces the fraud rate of the insured person and the loss of medical insurance funds on the basis of an intelligent technology, can realize the rapid large-scale application and replication of an anti-fraud scheme, provides anti-fraud safety services for more organizations and areas, improves the current single anti-fraud strategy, and promotes the anti-fraud strategy to evolve towards intellectualization and intellectualization.
3) The working efficiency of the mechanism is improved;
the traditional fraud detection method is realized through expert experience or an anti-fraud medical rule base, the anti-fraud of the strategy has strong regularity, and the strategy is easy to be drilled by a fraudster and has low efficiency; the medical insurance anti-fraud monitoring and analyzing system based on unsupervised isolated point detection not only can process batch data, but also can be applied to different scenes to enable fraud to be invisible. And the system model is adopted for judgment, so that the defects caused by insufficient experience or manpower are avoided, and the working efficiency is improved.
Drawings
FIG. 1 is a flow chart of a method in an embodiment of the invention.
Fig. 2 is an exemplary diagram of an isolated forest in the embodiment of the present invention.
FIG. 3 is a graph comparing normal point path segmentation and abnormal point path segmentation in the embodiment of the present invention.
FIG. 4 is a diagram illustrating an average path length in an embodiment of the invention.
FIG. 5 is a schematic diagram of evaluation indexes in the example of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
At present, a large amount of medical insurance funds in social insurance are drilled into vacancies by many lawbreakers, and the medical insurance funds are acquired by an unfair means, so that the medical insurance funds invisibly graze life-saving money of patients. These behaviors are not only bad, but also form a vicious circle if not remedied and managed, continuously eroding medical insurance funds, and increasing financial burden. Compared with the serious problem of cheating insurance, the current anti-cheating capability is insufficient, especially in a front-line city with relatively lagged technologies, the medical insurance wind control system is almost the original list class, and list information is seriously lacked. Or intercepting by using some simple rules, such as that one medical insurance card cannot be used for a doctor in 5 hospitals in one day, and the intercepting accuracy and coverage are very limited. The essence of the anti-fraud monitoring and analyzing solution for medical insurance based on unsupervised isolated point detection is that abnormal data is found and compared through technical means, detection results are given, and the abnormalities are hidden under model traversal.
The abnormal behaviors of the ginseng and insurance people refer to the diagnosis characteristics of the ginseng and insurance people by using the diagnosis behavior data generated in the diagnosis process of the ginseng and insurance people in the medical diagnosis process, samples different from most of the diagnosis characteristics of the ginseng and insurance people in the diagnosis characteristic samples of the ginseng and insurance people are called as the behavior noise of the ginseng and insurance people, and the suspected ginseng and insurance people can be determined by finding out the behavior noise of the ginseng and insurance people in the medical diagnosis process.
The abnormal behavior of the suspect is different from the definition of a general abnormal point or noise point, the abnormal behavior of the suspect is mainly the suspect, the diagnosis behavior data generated in the diagnosis and treatment process is used as the diagnosis characteristic, the behavior data in the diagnosis and treatment process is comprehensively considered, and then the abnormal value is found out by adopting the algorithm of the density degree of the data; however, the common abnormal value analysis or outlier identification is limited to the feature detection of a single feature or a one-dimensional data space, and the situation of comprehensively considering the multi-dimensional data features is lacked.
The embodiment of the invention firstly provides a medical insurance anti-fraud monitoring and analyzing method based on unsupervised isolated point detection, which comprises the following steps:
step S1, acquiring the visit behavior characteristic data of the ginseng insurance people to form an original sample set;
when acquiring the visit behavior characteristic data of the paramedics, classifying and summarizing characteristic variables related to medical insurance fraudulent behaviors, and mainly dividing the characteristic variables into hospital information, doctor information, patient information, visit item information, visit expense information and the like according to a data generation mechanism;
step S2, obtaining a plurality of visiting behavior characteristic data sets of the insured persons according to the original sample set;
for the original sample set, according to a sample number threshold, when the sample number is greater than the sample number threshold, the original sample set can be simply divided into a plurality of visit behavior characteristic data sets of the insured person; for example, 500 samples are collected in the original sample set, 5 visit behavior characteristic data sets of the insured person can be divided, and 100 samples are collected in each visit behavior characteristic data set of the insured person;
when the number of samples does not exceed the threshold value of the number of samples, a bagging method (namely Bootstrap) can be adopted to extract samples from the original sample set for multiple times to obtain k diagnosis behavior characteristic data sets of the insured person; k is more than or equal to 2; bootstrap is called a bagging method or a self-expanding method, and is a method for extracting and replacing, aiming at obtaining the distribution and confidence interval of statistics; the bagging method can combine a plurality of strong models which are better and more comprehensive if the classifier is expected to obtain;
step S3, creating a corresponding isolated tree according to the visit behavior feature data set of each insured person;
the path length of the sample points in the isolated tree is the number of edges which are passed by the sample points from the root node to the leaf node of the tree;
sample data set X ═ X1，x2……xnThe visit behavior characteristic data set of the insured person is obtained; wherein
xi＝(xi1,xi2,……，xid)；
S301, randomly extracting from X
A subset X' of X formed by sample points is placed into a root node;
s302, randomly appointing a dimension q from d dimensions, and randomly generating a segmentation point p, min (x) in the current subsetij,j＝q,xij∈X’)＜p＜max(xij,j＝q,xij∈X’)；
The dimension q is, for example, the visit fee, and the division point p is a specific numerical value of the visit fee, for example, 200 yuan;
s303, the dividing point p generates a hyperplane, and the current data space is divided into two subspaces: assigning sample points with the corresponding value of dimension smaller than p to be classified into a left child node, and assigning sample points with the value larger than or equal to p to be classified into a right child node;
s304, recursion S302 and S303, until all leaf nodes have only one sample point or the isolated tree has reached a specified height;
an isolated tree is established according to the steps, and the isolated tree has great randomness in judgment according to the process of establishing the tree; firstly, randomly selected variables and secondly randomly selected segmentation points; the randomness causes that the isolated tree hardly has strong robustness, and in order to solve the difficulty, a plurality of isolated trees are created by an integration method, and the generalization capability of the model is improved by utilizing a common decision mechanism;
step S4, integrating the isolated trees to form an isolated forest;
after the isolated forest exists, the model formed by each isolated tree can be subjected to average or voting to obtain the overall condition or overall evaluation;
an example of an isolated forest is shown in figure 2;
step S5, establishing an evaluation index for evaluating the visit behavior of the insured person based on the isolated forest;
as can be seen from fig. 3, the clusters with a high density need to be segmented many times to be isolated, but those with a low density can be easily isolated;
the comparison of the average path length of the outliers and the average path length of the normal points can be seen in fig. 4;
the evaluation indexes are established as follows:
for each sample xiTraversing each isolated tree of the isolated forest, and calculating the average path length of the isolated trees in the isolated forest
E (h (x)) is sample xiPath length expectation in solitary forests; average path length for all samples
Carrying out normalization processing; the calculation formula of the evaluation index is as follows:
wherein,
h (i) is a harmonic number, which can be estimated as ln (i) + 0.5772156649;
as can be seen from the view of figure 5,
when E (h (x)) → c (n), s → 0.5, it cannot be distinguished whether or not there is an abnormality; n is the number of samples;
when E (h (x)) → 0 and s → 1, that is, the abnormality score which is the evaluation index of the sample approaches 1, it is determined to be abnormal;
when E (h (x)) → n-1, s → 0, it is judged to be normal;
step S6, traversing each isolated tree of the isolated forest for the input visit behavior characteristic data sample of the insured person, and outputting a detection result according to the obtained evaluation index;
in the step, an evaluation index threshold value is specifically set, and when the obtained evaluation index is greater than the evaluation index threshold value, the abnormal behavior of the ginseng insurance person is judged; the corresponding ginseng protector is a suspected ginseng protector.
The embodiment of the invention also provides a medical insurance anti-fraud monitoring and analyzing system based on unsupervised isolated point detection, which comprises:
a memory storing a computer program;
a processor for executing the computer program, the computer program when executed performing the steps of the method as described above.
Finally, it should be noted that the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to examples, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (6)
1. A medical insurance anti-fraud monitoring and analyzing method based on unsupervised isolated point detection is characterized by comprising the following steps:
step S1, acquiring the visit behavior characteristic data of the ginseng insurance people to form an original sample set;
step S2, obtaining a plurality of visiting behavior characteristic data sets of the insured persons according to the original sample set;
step S3, creating a corresponding isolated tree according to the visit behavior feature data set of each insured person;
step S4, integrating the isolated trees to form an isolated forest;
step S5, establishing an evaluation index for evaluating the visit behavior of the insured person based on the isolated forest;
and step S6, traversing each isolated tree of the isolated forest for the input visit behavior characteristic data sample of the insured person, and outputting a detection result according to the obtained evaluation index.
2. The medical insurance anti-fraud monitoring and analyzing method based on unsupervised outlier detection of claim 1,
step S3 specifically includes:
sample data set X ═ X1，x2......xnThe visit behavior characteristic data set of the insured person is obtained; wherein
xi＝(xi1，xi2，......，xid)；
S301, randomly extracting from X
A subset X' of X formed by sample points is placed into a root node;
s302, randomly appointing a dimension q from d dimensions, and randomly generating a segmentation point p, min (x) in the current subsetij，j＝q，xij∈X’)＜p＜max(xij，j＝q，xij∈X’)；
S303, the dividing point p generates a hyperplane, and the current data space is divided into two subspaces: assigning sample points with the corresponding value of dimension smaller than p to be classified into a left child node, and assigning sample points with the value larger than or equal to p to be classified into a right child node;
s304, recursion S302 and S303, until all leaf nodes have only one sample point or the isolated tree has reached a specified height.
3. The medical insurance anti-fraud monitoring and analyzing method based on unsupervised outlier detection of claim 2,
the evaluation indexes are established as follows:
for each sample xiTraversing each isolated tree of the isolated forest, and calculating the average path length of the isolated trees in the isolated forest
E (h (x)) is sample xiPath length expectation in solitary forests; average path length for all samples
Carrying out normalization processing; the calculation formula of the evaluation index is as follows:
wherein,
h (i) is a harmonic number.
4. The medical insurance anti-fraud monitoring and analyzing method based on unsupervised outlier detection of claim 3,
in step S6, an evaluation index threshold is specifically set, and when the obtained evaluation index is greater than the evaluation index threshold, it is determined that the person is an abnormal behavior of the insured person; the corresponding ginseng protector is a suspected ginseng protector.
5. The method for medical insurance anti-fraud monitoring and analysis based on unsupervised outlier detection according to any of claims 1-4,
step S2 specifically includes:
for an original sample set, when the number of samples is larger than a sample number threshold value, simply dividing the original sample set into a plurality of visit behavior characteristic data sets of the insured persons;
when the number of samples does not exceed the threshold value of the number of samples, extracting the samples from the original sample set for multiple times by adopting a bagging method to obtain k diagnosis behavior characteristic data sets of the insured person; k is more than or equal to 2.
6. A medical insurance anti-fraud monitoring and analyzing system based on unsupervised isolated point detection is characterized by comprising:
a memory storing a computer program;
a processor for running the computer program, the computer program when running performing the steps of the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010868057.9A CN111951116A (en) | 2020-08-26 | 2020-08-26 | Medical insurance anti-fraud monitoring and analyzing method and system based on unsupervised isolated point detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010868057.9A CN111951116A (en) | 2020-08-26 | 2020-08-26 | Medical insurance anti-fraud monitoring and analyzing method and system based on unsupervised isolated point detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111951116A true CN111951116A (en) | 2020-11-17 |
Family
ID=73367869
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010868057.9A Withdrawn CN111951116A (en) | 2020-08-26 | 2020-08-26 | Medical insurance anti-fraud monitoring and analyzing method and system based on unsupervised isolated point detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111951116A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022142042A1 (en) * | 2020-12-29 | 2022-07-07 | 平安科技（深圳）有限公司 | Abnormal data detection method and apparatus, computer device and storage medium |
| CN114881775A (en) * | 2022-07-12 | 2022-08-09 | 浙江君同智能科技有限责任公司 | Fraud detection method and system based on semi-supervised ensemble learning |
| CN117874653A (en) * | 2024-03-11 | 2024-04-12 | 武汉佳华创新电气有限公司 | A method and system for power system security monitoring based on multi-source data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150046181A1 (en) * | 2014-02-14 | 2015-02-12 | Brighterion, Inc. | Healthcare fraud protection and management |
| CN109376381A (en) * | 2018-09-10 | 2019-02-22 | 平安科技（深圳）有限公司 | Method for detecting abnormality, device, computer equipment and storage medium are submitted an expense account in medical insurance |
| CN109976930A (en) * | 2017-12-28 | 2019-07-05 | 腾讯科技（深圳）有限公司 | Detection method, system and the storage medium of abnormal data |
| CN110189232A (en) * | 2019-05-14 | 2019-08-30 | 三峡大学 | Abnormal Analysis Method of Electricity Information Collection Data Based on Isolated Forest Algorithm |
-
2020
- 2020-08-26 CN CN202010868057.9A patent/CN111951116A/en not_active Withdrawn
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150046181A1 (en) * | 2014-02-14 | 2015-02-12 | Brighterion, Inc. | Healthcare fraud protection and management |
| CN109976930A (en) * | 2017-12-28 | 2019-07-05 | 腾讯科技（深圳）有限公司 | Detection method, system and the storage medium of abnormal data |
| CN109376381A (en) * | 2018-09-10 | 2019-02-22 | 平安科技（深圳）有限公司 | Method for detecting abnormality, device, computer equipment and storage medium are submitted an expense account in medical insurance |
| CN110189232A (en) * | 2019-05-14 | 2019-08-30 | 三峡大学 | Abnormal Analysis Method of Electricity Information Collection Data Based on Isolated Forest Algorithm |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022142042A1 (en) * | 2020-12-29 | 2022-07-07 | 平安科技（深圳）有限公司 | Abnormal data detection method and apparatus, computer device and storage medium |
| CN114881775A (en) * | 2022-07-12 | 2022-08-09 | 浙江君同智能科技有限责任公司 | Fraud detection method and system based on semi-supervised ensemble learning |
| CN117874653A (en) * | 2024-03-11 | 2024-04-12 | 武汉佳华创新电气有限公司 | A method and system for power system security monitoring based on multi-source data |
| CN117874653B (en) * | 2024-03-11 | 2024-05-31 | 武汉佳华创新电气有限公司 | Power system safety monitoring method and system based on multi-source data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10872131B2 (en) | Progression analytics system | |
| Anitha et al. | Development of computer‐aided approach for brain tumor detection using random forest classifier | |
| CN112200684B (en) | A method, system and storage medium for detecting medical insurance fraud | |
| CN112989332B (en) | Abnormal user behavior detection method and device | |
| WO2019019630A1 (en) | Anti-fraud identification method, storage medium, server carrying ping an brain and device | |
| US20160110512A1 (en) | Method of personalizing, individualizing, and automating the management of healthcare fraud-waste-abuse to unique individual healthcare providers | |
| CN112991079B (en) | Multi-card co-occurrence medical treatment fraud detection method, system, cloud end and medium | |
| Sun et al. | Patient cluster divergence based healthcare insurance fraudster detection | |
| CN111951116A (en) | Medical insurance anti-fraud monitoring and analyzing method and system based on unsupervised isolated point detection | |
| CN113095365A (en) | Medical insurance violation data identification method and device | |
| Lokanan | The determinants of investment fraud: A machine learning and artificial intelligence approach | |
| CN113642672A (en) | Feature processing method and device of medical insurance data, computer equipment and storage medium | |
| Gaurav et al. | Bankruptcy forecasting in enterprises and its security using hybrid deep learning models | |
| Wanke et al. | Performance evaluation and lockdown decisions of the UK healthcare system in dealing with COVID-19: A novel unbiased MCDM score decomposition into latent vagueness and randomness components | |
| Cheng et al. | Research on medical insurance anti-gang fraud model based on the knowledge graph | |
| CN109635112A (en) | Abnormal dialysis data screening method, apparatus, equipment and storage medium | |
| CN116865994A (en) | Network data security prediction method based on big data | |
| Becker et al. | Rough set theory in the classification of loan applications | |
| CN118737470B (en) | Financial data verification method and system | |
| CN115545955A (en) | Abnormal data detection method, device and electronic equipment in medical record data | |
| Cattinelli et al. | Computational intelligence for the Balanced Scorecard: Studying performance trends of hemodialysis clinics | |
| CN117271508A (en) | Detection method and system for abnormal aggregation of medical insurance cards | |
| CN114493899A (en) | Method and system for constructing classification prediction model of authenticable state | |
| Sakai et al. | Healthcare fraud detection using data mining | |
| Malvar et al. | Machine learning approaches for localized lockdown during covid-19: a case study analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20201117 |